U.S. patent number 6,055,007 [Application Number 08/219,445] was granted by the patent office on 2000-04-25 for image processing device and image reading device.
This patent grant is currently assigned to Minolta Co., Ltd.. Invention is credited to Yoko Fujiwara, Takashi Honda, Masaaki Kuriyama.
United States Patent |
6,055,007 |
Fujiwara , et al. |
April 25, 2000 |
Image processing device and image reading device
Abstract
This invention concerns an image processing device for adjusting
the density of an image and an image reading device incorporating
the image processing device and aims to enhance the repeatability
of pseudo-gradient without reference to density distribution of the
image. The image processing device which fulfills the purpose of
compensating the gradient of image data consisting of multivalue
digital signals comprises means ME1 to ME3 for storing a plurality
of gradient compensation data T1 to T3 having mutually different
contents of compensation, means for detecting a contrast of an
image on the basis of a group of image data conforming to one
image, and means for switching the contents of compensation for the
image data by selecting one of the three gradient compensation data
T1 to T3 in conformity with the contrast.
Inventors: |
Fujiwara; Yoko (Zama,
JP), Kuriyama; Masaaki (Machida, JP),
Honda; Takashi (Machida, JP) |
Assignee: |
Minolta Co., Ltd. (Osaka,
JP)
|
Family
ID: |
13518360 |
Appl.
No.: |
08/219,445 |
Filed: |
March 29, 1994 |
Foreign Application Priority Data
|
|
|
|
|
Mar 31, 1993 [JP] |
|
|
5-073442 |
|
Current U.S.
Class: |
347/131 |
Current CPC
Class: |
H04N
1/40056 (20130101); H04N 1/4074 (20130101) |
Current International
Class: |
H04N
1/40 (20060101); H04N 1/407 (20060101); B41J
002/435 (); G01D 015/14 (); H01J 029/70 (); H01J
033/00 () |
Field of
Search: |
;318/96 ;250/214,571
;347/131 ;358/443,448,447,296,298 |
References Cited
[Referenced By]
U.S. Patent Documents
Foreign Patent Documents
|
|
|
|
|
|
|
63-215262A |
|
Sep 1988 |
|
JP |
|
4-37258A |
|
Feb 1992 |
|
JP |
|
Primary Examiner: Barlow; John
Assistant Examiner: Gordon; Raquel Yvette
Attorney, Agent or Firm: Burns, Doane, Swecker & Mathis,
LLP
Claims
What is claimed is:
1. An image processing device for processing multivalue image data,
comprising:
a memory having stored therein a plurality of different sets of
tone compensation data;
detecting means for detecting a density contrast of an image
represented by a plurality of image data;
selecting means for selecting one of said sets of tone compensation
data in response to the density contrast detected by said detecting
means; and
compensating means for compensating a tone of the image data based
on the tone compensation data selected by said selecting means.
2. An image processing device according to claim 1, which further
comprises means for forming a histogram of numerical values of said
plurality of image data and wherein said detecting means detects
said density contrast based on a width of said histogram.
3. An image processing device according to claim 1, which further
comprises means for causing the image data having a gradient
thereof compensated by said compensating means to be binarized
based on a threshold.
4. An image processing device according to claim 3, which further
comprises means for forming a histogram of the numerical values of
said group of image data and threshold setting means for setting
said threshold on the basis of said histogram.
5. An image processing device according to claim 4, wherein said
threshold setting means, when said histogram has two peaks, for
setting said threshold at a magnitude equal to a median of data
values corresponding to said peaks.
6. An image processing device according to claim 1, wherein said
tone compensation data are in a form of a lookup table showing
correspondence between values of image data for input and the
values of the image data for output.
7. An in age reading device, comprising:
a lamp;
an image sensor for reading an image on an original document
illuminated with said lamp and issuing multivalue image data in
conformity with a density of each pixels of said image;
a memory for storing a plurality of gradient compensation data
having mutually different contents;
detecting means for detecting a contrast of an image indicated by a
plurality of the image data produced by said image sensor;
selecting means for selecting one of said gradient compensation
data in conformity with the contrast detected by said detecting
means;
compensating means for compensating the gradient of said image data
based on the gradient compensation data selected by said selecting
means;
adjusting means for adjusting luminous energy of said lamp so that
a median in a substantial range of compensation based on the
gradient compensation data selected by said selecting means
coincides with a threshold; and
means for causing the image data having the gradient thereof
compensated by said compensating means to be binarized on the basis
of said threshold.
8. An image reading device according to claim 7, which further
comprises means for forming a histogram of values of said plurality
of image data produced by said image sensor and wherein said
detecting means detects said contrast on the basis of a width of
said histogram.
9. An image reading device according to claim 7, which further
comprises means for forming a histogram of the values of said group
of image data produced from said image sensor and threshold setting
means for setting said threshold on the basis of said
histogram.
10. An image reading device according to claim 9, wherein said
threshold setting means, when said histogram has two peaks, sets
said threshold at a magnitude equal to the median of data values
corresponding to said peaks.
11. An image reading device according to claim 7, which further
comprises control means for controlling said image sensor so as to
read the image of said original document again after the adjustment
by said adjusting means and wherein said compensating means
compensates a tone of the image data produced from said image
sensor after the adjustment by said adjusting means.
12. An image reading device according to claim 11, wherein said
binarizing means binarizes the image data produced by said
compensating means after the adjustment by said adjusting means is
effected.
13. A method for reading an image by use of an image reading device
provided with a memory for storing a plurality of gradient
compensation data having mutually different contents, comprising
the steps of:
illuminating an original document with a lamp;
reading the image on said original document illuminated by said
lamp and producing multivalue image data conforming with a density
of each pixels of said image;
detecting a contrast of image indicated by a group of produced
image data;
selecting one of said gradient compensation data in conformity with
said detected contrast;
adjusting luminous energy of said lamp so that a median in a
substantial range of compensation based on the gradient
compensation data selected by said selecting means coincides with a
threshold;
reading the image on said original document again and producing
multivalue image data conforming with the density of each of said
pixels;
compensating the gradient of the produced image data based on the
selected gradient compensation data; and
binarizing the image data having the gradient thereof compensated
based on said threshold.
Description
BACKGROUND OF THE INVENTION
This invention relates to an image processing device for adjusting
the density of an image and an image reading device having the
image processing device incorporated therein.
As image input means for such digital devices as copying machine,
facsimile, film scanner, OCR (optical character reader), optical
filing system, and various display devices which form, memorize,
and display images by the combination of dots (pixels) of a fixed
size, an image reading device (image reader) which reads an image
on a given original document in the form of an aggregate of
distinctly divided pixels has been in popular use.
Generally, the image reading device is so constructed as to operate
by a procedure which comprises reading the image of a given
original document by the use of a one-dimensional image sensor
extending in the direction of main scanning, converting the analog
signals of the individual pixels obtained by the reading into pixel
data in the form of multivalue digital signals, subjecting the
pixel data resulting from the conversion to image processing, and
producing binary image data corresponding to the individual
pixels.
The image processing which is performed in the image reading device
is known in numerous forms including edge enhancing and smoothing
processings for the improvement of image quality, trimming and
negative-positive inverting processings for the editing of an
image, density conversion processings for effecting tone
compensation in accordance with the .gamma. characteristics of an
externally connected device (image reproducing means such as a
printer) to intensify the density gradient during the reproduction
of an image, and binarization processings for the conversion of
multivalue data in accordance with the density of each of the
pixels of an original document into binary data, for example.
The so-called lookup table method is widely adopted for the density
conversion processings. To be specific, the compensation of image
data is accomplished by installing an ROM (read only memory) having
stored therein a conversion table which is a collection of
necessary compensation image data and consequently enabling such
compensation image data as are stored at addresses answering the
numerical values of input image data to be read out in the form of
output image data D.
With respect to binarization processing, the so-called simple
binarization processing which resides in comparing a multivalue
level of one given pixel with a fixed threshold level is suitable
for handling such linear images as characters from the point of
view of the quality of a reproduced image and the pseudogradient
processing which resorts to the dither method or the error
diffusion method is suitable for handling such intermediate tone
images as photographs. Where an original document which contains
both linear images and intermediate tone images in a mixed state is
to be read, therefore, the simple binarization processing and the
pseudogradient processing are used as switched in accordance with
the attributes of a given image which decide between a linear image
and an intermediate tone image (hereinafter referred to
occasionally as "binarization attributes").
Incidentally, the conventional image reading device has performed
the density conversion processing by the use of one and the same
conversion table. In other words, the contents of the tone
compensation performed on image data have been set and fixed so as
to exalt the reproducibility of images relative to images of
standard density distribution.
Thus, in the binarization by the pseudogradient processing, the
conventional image reading device suffers from impairment of the
reproducibility of white and black parts of an image having a large
density contrast and conversely entails the problem of reproducing
such an unnatural image as a binary image from an image having a
small density contrast.
SUMMARY OF THE INVENTION
This invention has for its object the exaltation of the
reproducibility of pseudogradient processing without reference to
the density distribution of an image.
To accomplish the object described above, this invention provides
an image processing device for processing multivalue image data,
comprising a memory having stored therein a plurality of sets of
tone compensation data having mutually different contents,
detecting means for detecting a density contrast of an image
represented by a group of image data, selecting means for selecting
one of the aforementioned sets of tone compensation data in
response to the density contrast detected by the detecting means,
and compensating means for compensating the tone of image data on
the basis of the tone compensation data selected by the selecting
means.
To accomplish this object, the present invention further provides
an image reading device, comprising a lamp, an image sensor for
reading an image on an original document illuminated with the lamp
and producing multivalue image data corresponding to the density of
each of the individual pixels of the image, compensating means for
compensating the tone of image data produced from the image sensor
on the basis of tone compensation data, adjusting means for
adjusting the luminous energy of the lamp so as to establish
concurrence between the median in the substantial range of
compensation due to the tone compensation data and a threshold, and
binarization means for binarizing the image data having the tone
thereof compensated by the aforementioned compensating means on the
basis of the aforementioned threshold.
BRIEF DESCRIPTION OF THE DRAWINGS
FIG. 1 is a sectioned front view illustrating schematically the
construction of an image reader according to this invention.
FIG. 2 is a block diagram of an electric circuit part of the image
reader of FIG. 1.
FIG. 3 is a diagram illustrating the construction of a density
converting part of FIG. 2.
FIG. 4 is a block diagram of an image processing part of FIG.
2.
FIG. 5 is a main flow chart illustrating schematically the
operation of a CPU.
FIG. 6 is a flow chart illustrating reception processing.
FIG. 7 is a flow chart illustrating transmission processing.
FIG. 8 is a flow chart illustrating the initial setting processing
of FIG. 5.
FIG. 9 is a flow chart illustrating the reading processing of FIG.
5.
FIG. 10 is a flow chart illustrating the read mode designating
processing of FIG. 5.
FIG. 11 is a flow chart illustrating the density designating
processing of FIG. 10.
FIG. 12 is a flow chart illustrating the attributes designating
processing of FIG. 5.
FIG. 13 is a flow chart illustrating the attributes data writing
processing of FIG. 12.
FIG. 14 is a flow chart illustrating the output data designating
processing of FIG. 5.
FIG. 15 is a histogram showing one example of the density
distribution of an image.
FIG. 16 is a graph showing the input/output characteristics of a
density conversion part as the function of the luminous energy used
for scanning.
FIG. 17 is a graph showing the change in the input/output
characteristics of a density conversion part caused by switch of
conversion tables.
FIG. 18 is a graph showing a typical alteration caused in the
contents of density variation by adjustment of the luminous energy
used for scanning.
FIG. 19 is a histogram showing one example of the density
distribution of a character image.
DESCRIPTION OF THE PREFERRED EMBODIMENTS
FIG. 1 is a sectioned front view illustrating schematically the
construction of an image reader according to this invention.
An image reader 1 is provided on the upper surface of a rectangular
housing with a platen glass 2 which is capable of supporting
thereon an original document of the A3 size and an original
document cover 3 adapted to keep an original document placed on the
platen glass 2 fast in place. The platen glass 2 is provided at the
front end part thereof relative to the direction of scanning with a
standard pattern 14 which is formed of a white image for the
compensation of shading.
Inside the housing, there are provided an optical system set in
place so as to be capable of scanning an image on the original
document in the direction of an arrow M5 (direction of
sub-scanning) below the platen glass 2 and an electric circuit
system 12 adapted to form image data in conformity with the density
or color of the image of the original document.
The optical system is composed of a first slider 13 incorporating
an illumination lamp 4, a reflection mirror 5, and a mirror 6
therein, a second slider 13a incorporating a mirror 7 and a mirror
8 therein, and a main lens 9, etc. The first slider 13 and the
second slider 13a are driven as controlled so that the rate of
motion of the second slider 13a is v/2 relative to the rate of
motion of the first slider 13 taken as v.
The scanning light which has passed through the main lens 9 is
allowed to impinge on an image sensor 10 attached to a supporting
member 11 and is converted into an electric signal (image signal).
The image sensor 10 is formed of a plurality of CCD chips arranged
in a continuous pattern in the direction of main scanning
(direction of line) and adapted to read the original document at a
resolution of 400 pixels/inch. In each of the CCD chips, a
multiplicity of light-receiving elements are arranged in a row. The
light-receiving elements in the row are divided into three regions.
The component light-receiving elements of these three regions are
severally provided on their surfaces with spectral filters such
that the regions severally admit the three colors, R (red), G
(green), and B (blue). The light-receiving elements correspond one
each to the pixels of the image on the original document and they
severally emit to the electric circuit part 12 an image signal
conforming with the intensity of a reflected light relative to one
of the colors of relevant pixels.
FIG. 2 is a block diagram of the electric circuit part 12 of the
image reader 1.
The electric circuit part 12 is composed of a color separation part
21 for separating the image signal from the image sensor 10 into
signals of the colors R, G, and B and subjecting the separated
color signals to prescribed amplification, a digitization
processing part 22 for quantizing the analog signals in different
colors and emitting image data D7 to D0 of eight bits (256
gradients), a shading compensation part 23 for compensating the
dispersion in the direction of main scanning of the image data D7
to D0 due to an uneven distribution of the light from the
illuminating lamp 4 or a difference in sensitivity among the bits
of the image sensor 10, an image discriminating part 25 for
discriminating the attributes of binarization and the attributes of
color, a lookup table type density conversion part (.gamma.
conversion ROM) 26 for performing .gamma. conversion in conformity
to the density level adjustment and the density characteristics
(.gamma. characteristics) of the externally connected device, an
image processing part 28 for performing digital processing
including binarization and image editing, an output control part 29
for controlling data output, an attribute memory 30 for recording
designated attribute data a2 to a0, an attribute data output
control part 31, a clock generation circuit 41, a line memory 24
for memorizing one line full of the image data D17 to D10 issued
from the shading compensation part 23, a synchronous signal
generation part 40 for issuing various synchronization signals, a
lamp control part 4a for controlling the switching of the
illumination lamp 4, a driver 16a for driving a scanner motor 16
for scanning, and a CPU 20 for controlling all of the component
parts mentioned above.
The CPU 20 has built therein a ROM 201 for storing a processing
program, a register 202 for storing various flags and status data
temporarily therein in execution of the program, and a RAM 203
destined to form a work area. The CPU 20 performs communication for
the reception and transmission of data indicative of a varying
command and the operation state (status) of the image reader 1 with
an external host device fitted with manual operation means, forms
designation attribute data a2 to a0 for designating image editing
or binarization processing based on the received commands prior to
the reading of the image of an original document, and stores the
designation attribute data a2 to a0 in the attribute memory 30. In
the image reader 1, the designation attribute data a0, a1, and a2
are the data for designating the binarization processing, the
negative-positive inversion, and the trimming respectively.
Incidentally, the synchronization signals to be issued by the
synchronization signal generator 40 include a horizontal
synchronization signal Hsync which is issued for each line of main
scanning, an pixel clock signal SYNCK which forms the standard for
data transmission timing for each pixel, and an output enable
signal VD which indicates the effective duration for the data
issued from the image reader 1, for example. In the shading
compensation part 23, the image data D7 to D0 are subjected to
shading compensation and, at the same time, these image data D7 to
D0 which have been data proportional to the intensity of the
reflected light are converted by a logarithmic computation based on
visional characteristics into density data proportional to the
density of an image.
The image discrimination part 25 is intended to form a judgment to
discriminate between a character region and a photograph region in
an image or decide whether or not the color edition has a place
designated therefor. The discrimination attribute data .alpha.0
issued by the image discrimination part 25 assumes the value "0"
when the divided region subjected to the discrimination corresponds
to a character image (character region) or the value "1" when this
divided region corresponds to an intermediate tone image
(photograph region). In contrast thereto, the discrimination
attribute data a1 assumes the value "1" when the divided region
subjected to the discrimination corresponds to a specific color or
the value "0" when the divided region corresponds to a color other
than the specific color.
FIG. 3 is a block diagram illustrating the construction of the
density conversion part 26.
The density conversion part 26 is formed of a .gamma. conversion
ROM incorporating three memory regions (banks) ME1 to ME3 therein.
The memory regions ME1 to ME3 have respectively stored therein
conversion tables T1 to T3 for input/output characteristics which
will be described specifically afterward.
The conversion tables T1 to T3 are selectively used by the density
conversion control which is effected by the CPU 20. To be specific,
in the .gamma. conversion, the CPU 20 allows access to one of the
memory regions ME1 to ME3 and causes the data at an address
designated by the memory region and the value "0" to "255" of the
image data D17 to D10 received from the shading compensation part
23 to be read out as image data D27 to D20.
FIG. 4 is a block diagram illustrating the image processing part
28.
The image processing part 28 is intended for the processing of an
image signal introduced from the image sensor 10 and, thus, is
composed of a variable power part 281, a filtering part 282, a
trimming-masking part 283, a simple binarization processing part
284, a pseudo-intermediate tone processing part 285, a data
selection part 286, a selector 287, and a negative processing part
288.
To the image processing part 28, the image data D27 to D20 from the
density conversion part 26 are serially injected in the order of
arrangement of pixels. The image data D27 to D20 thus introduced
are first subjected in the variable power part 281 to variable
power processing as set by the CPU 20. Then, in the filtering part
282, the data are subjected to processings for the improvement of
image quality such as edge enhancement and smoothing to give rise
to image data D37 to D30 to be discharged.
In the trimming-masking part 283 which operates in response to the
designation attribute data a2, when the data a2 is "1", the
processing of masking forces the image data D37 to D30 to assume
the value of "0" corresponding to a blank part and to be
consequently discharged as image data D47 to D40. When the data a2
is "0", the image data D37 to D30 are passed unaltered and are
discharged as image data D47 to D40 (data through).
The image data D47 to D40 issued from the trimming-masking part 283
are binarized in the simple binarization processing part 284 and
the pseudo-intermediate tone processing part 285 and are
simultaneously discharged as binary image data DA, DB to be
injected as such into the selector 287.
The selector 287 selects either of the two binary image data DA, DB
in accordance with the output data D-SEL from the data selection
part 286 and emits the selected data as an output. To the data
selection part 286, the designation attribute data a0 for
controlling the binarization processing are imparted together with
the aforementioned discrimination attribute data .alpha.0 obtained
by automatic discrimination of the binary attribute. The value of
the output data D-SEL is fixed by the value of the data a0. To be
specific, when the data a0 is "0", the discrimination attribute
data .alpha.0 are emitted unaltered as the output data D-SEL. When
the data a0 is "1", the data obtained by the inversion of the
discrimination attribute data .alpha.0 are issued as an output.
In the image processing part 28, when the designation attribute
data a0 is "0", the external designation of the binarization
processing is defaulted and the binary image data based on the
automatic discrimination of the binarization attribute by the image
discrimination part 25 are issued as an output. When the data a0 is
"1", the binary image data which have undergone the binarization
processing which is the opposite of the result of the automatic
discrimination of the binarization attribute are issued as an
output.
The negative processing part 288 issues the binary image data
introduced from the selector 287 in their unaltered form when the
designation attribute data a1 is "0" or the inverted binary image
data when the data a1 is "1" respectively as image data VIDEO 0 to
7 each composed of eight parallel pixels.
Now, the operation of the image reader 1 will be described below
with reference to the flow charts of FIGS. 5 to 14.
FIG. 5 is a main flow chart schematically illustrating the
operation of the CPU 20.
When the power source is connected to the system and the program is
started, the system is initialized at Step #1.
Then, the presence or absence of a command from the host device is
checked at Step #2. When the presence of a command is confirmed,
the kind of this command is discerned (Step #3) and, depending on
the kind of the command, the read processing (Step #4), the read
mode designation processing (Step #5), the attribute designation
processing (Step #6), and the output data designation processing
(Step #7) are executed.
Thereafter, other processings such as for effecting detection of
the status of operation (Step #8) are executed and the program is
returned to Step #2. The processings at Steps #2 to #8 are
subsequently repeated.
FIG. 6 is a flow chart illustrating the reception processing and
FIG. 7 is a flow chart illustrating the transmission
processing.
These routines are interrupt routines and are executed from time to
time in response to an access from the host device.
In the reception processing of FIG. 6, first the received signal is
subjected to code analysis (Step #11). When the reception of a
command is confirmed at Step #12, the received command is stored at
a prescribed area in the register 202 (Step #13).
When the received signal happens to designate a request for
information on the status (Step #14), the data indicating the
status such as the state of wait are read out of the register 202
and transmitted to the host device (Step #15).
When the received signal does not correspond to either the
predefined command or the request for status, code data which
indicate a reception error are transmitted (Step #16).
In the transmission processing shown in FIG. 7, the system waits
until the previous transmission is completed and the next
transmission is readied (Step #21) and code data to be transmitted
are set in the register 202 (Step #22). Then, at Step #23, the
presence or absence of code data to be subsequently transmitted,
namely the presence or absence of necessity for transmission, is
checked. When the necessity for transmission is confirmed, the flow
of processing is returned to Step #21.
FIG. 8 is a flow chart of the initialization processing at Step #1
shown in FIG. 5.
First, the status "WAIT" which indicates that the system is being
readied for read scanning is set. Specifically, the data
corresponding to "WAIT" are stored at the status area in the
register 202 (Step #31).
Then, at Step #32, a self test for checking the system to find
whether or not the component parts are normally operating is
carried out. At Step #33, the presence or absence of a mechanical
trouble in the system is checked. When the presence of a mechanical
trouble is confirmed, the flow of processing is moved to Step #37
to effect transmission of the code inhibiting operation to the host
device. When the absence of a mechanical trouble is confirmed, the
flow of processing is advanced to Step #34 to effect initialization
of the relevant component parts.
In the initialization at this time, "0" is written in the attribute
memory 30 as the designation attribute data a0, a1, and a2.
Thereafter in the image processing part 28, therefore, the image
editing processings by trimming and negative-positive inversion are
not executed and the binarization processing is destined to be
carried out on the basis of the discrimination attribute data
.alpha.0 unless the designation attribute data a2 to a0 are not
rewritten. The density is set at a standard level in the density
conversion part 26 and the inlet to the selector is so selected in
the output control part 29 that the image data VIDEO 0 to 7 and the
attribute data a2 to a0 are alternately issued.
After the initialization which is carried out as described above,
the first slider 13 is moved to the home position (Step #35) and,
subsequent to this motion, the status is changed from "WAIT" to
"READY" representing the state of readiness (Step #36).
FIG. 9 is a flow chart illustrating the read processing performed
at Step #4 as shown in FIG. 5.
First, the status is set at "BUSY" which indicates that the reading
is in process (Step #41) and the illumination lamp 4 is turned on
(Step #42).
Then, a scanner motor 16 is turned on (Step #43) and the arrival of
the first slider 13 at the shading position, namely a position
directly below the standard pattern 14, is waited (Step #44).
After the arrival of the slider 13 at the standard pattern 14, the
standard pattern 14 is read for the sake of compensation of the
shading and the standard image data (white data) are stored in the
line memory 24 (Step #45).
Subsequently, the arrival of the slider 13 at the leading end
position of the original document is waited at Step #46 and the
synchronizing signal generation part 40 is turned on and made to
issue a signal to synchronize at Step #47. As a result, the
relevant parts of the system operate in response to the signal to
synchronize and the image data VIDEO 0 to 7 and the attribute data
a4 to a0 which are rendered effective after the scanning of the 9th
line is started are alternately issued.
The completion of the scanning of the whole image on the original
document, namely the arrival of the slider 13 at the trailing end
position of the original document, is waited (Step #48). Then, the
synchronizing signal generation part 40 is turned off (Step #49),
the scanner motor 16 is provisionally turned off (Step #50), and
the illumination lamp 4 is turned off (Step #51).
Now, the scanner motor 16 is operated reversely to set the sliders
13, 13a in a return motion (Step #52), the return of the slider 13
to the home position is waited (Step #53), the scanner motor 16 is
turned off (Step #54), and finally the status is set at "READY" in
Step #55.
FIG. 10 is a flow chart illustrating the read mode designation
processing which is performed at Step #5 in the flow of processing
shown in FIG. 5.
The status is set at "WAIT" at Step #61, the parameter contained in
the command is checked at Step #62, in accordance with the
parameter, the density is designated (Step #63), the ratio of
variable power is designated (Step #64), and other designations
such as the designation of a device as the destination of the
output are executed (Step #65). Then, the status is returned to
"READY" at Step #66.
FIG. 11 is a flow chart illustrating the density designation
processing which is performed at Step #63 in the flow of processing
shown in FIG. 10.
First, a preliminary scanning for detecting the trend of image
density is carried out and the image data D17 to D10 which are
sequentially stored in the line memory 24 are admitted from time to
time in the RAM 203 at Step #71. In the preliminary scanning, the
sliders 13, 13a are moved at a higher rate than in the main
scanning to read the original document at a coarse pitch of 2 mm,
for example. In this case, the luminous energy of the illuminating
lamp 4 is set at a level at which the output of the CCD to be
emitted when the CCD reads out the standard image (white image)
verges on saturation.
At Step #72, the outputs (data values) of the component pixels are
produced based on the image data D17 to D10 of the RAM 203 and are
plotted to obtain a histogram. In this case, since the
density-output characteristics existing under the conditions of
preliminary scanning (such as the luminous energy of the lamp, the
sensitivity of the image sensor 10, and the AD conversion
characteristics) are already known, this histogram can be converted
into a histogram showing the density of each of the component
pixels as shown in FIG. 15. The new histogram affords data of
density distribution.
Then, the image contrast C is obtained on the basis of the
histogram of density at Step #73. Though the contrast C ought to
represent theoretically the difference between the maximum and the
minimum of density on the original document, it is actually
obtained for the sake of the present invention on the basis of the
maximum and the minimum of a range to be fixed by deducting several
% from each of the opposite ends of the total range of density in
due consideration of the effects of electrical noise and dust dust
on the original document.
At Step #74, one of the three memory regions (banks) ME1 to ME3 of
the density conversion part 26 (hereinafter referred to as ".gamma.
conversion ROM") is selected as specifically described afterward in
accordance with the contract C obtained above. The banks ME1 to ME3
of the .gamma. conversion ROM 26 have severally stored therein
conversion tables T1 to T3 which respectively correspond to three
density-output characteristics (.gamma. curves). The three
density-output characteristics of the present embodiment (herein
referred to respectively as .gamma. curve 1, .gamma. curve 2, and
.gamma. curve 3) are invariably such that the output data values
are proportional to the densities within a prescribed range, though
the lines which depict the relation under discussion vary from one
.gamma. curve to another. In other words, they are the ranges of
input in which the output values are varied with the input-output
characteristics, namely the ranges of density (substantial ranges
of compensation) .gamma.1, .gamma.2, and .gamma.3, in which the
gradients are repeatable during the pseudo-intermediate tone
processing by the dither method, are different from one another
(.gamma.1<.gamma.2<.gamma.3).
When the value of the contrast C is smaller than (.gamma.1
+.delta.), the conversion table T1 of the bank ME1 is selected as
the lookup table for the purpose of the .gamma. conversion. The
symbol .delta. stands for an empirically optimized constant. The
conversion table T2 of the bank ME2 is selected when the value of
the contrast C is an intermediate between (.gamma.1+.delta.) and
(.gamma.2+.delta.) and the conversion table T3 of the bank ME3 is
selected when the value of the contrast C is larger than
(.gamma.2+.delta.).
In the repetition of the gradient of a photographic image, for
example, an appropriate pseudo-intermediate tone image can be
obtained in spite of the contrast C by using a conversion table of
a .gamma. curve of a large inclination for an image of a small
contrast C or a conversion table of a .gamma. curve of a small
inclination for an image of a large contrast C as described above.
In short, where the .gamma. curve of the .gamma. conversion is
fixed as has been conventionally usual, the repeatability of the
white and black parts is impaired by extremely increasing the
contrast C as compared with the range of density in which the
gradient is repeatable and, conversely, such an unnatural image as
a binary image is obtained by extremely decreasing the contrast C.
The repeatability of gradient is improved by altering the .gamma.
curve of the .gamma. conversion proportionately to the contrast
C.
Then, at Step #75, the method of setting the density is checked to
decide whether or not the setting is automatically effected. When
the method of setting is not automatic, namely when it is manual,
the luminous energy of the lamp is set on the basis of an operation
designated by an operator at Step #76. The value of density
designated by the operator corresponds to the image density which
forms the threshold of the simple binarization (hereinafter
referred to a "threshold density Nth"). In the positive image, the
threshold density Nth increases in proportion as the value of
designated density increases. Here, the action of changing the
luminous energy of the lamp in conformity with the value of
designated density equals the action of fixing the luminous energy
of the lamp and altering the threshold density Nth by shifting the
.gamma. curve as illustrated in FIG. 16. In contrast thereto, the
action of changing the inclination of the .gamma. curve
proportionately to the contrast C in the case of a specific value
of designated density as described above equals the action of
suitably selecting the ranges of density .gamma.1, .gamma.2, and
.gamma.3 permitting repetition of gradient as illustrated in FIG.
17.
At Step #76, the luminous energy of the lamp is set so that in the
.gamma. curve of a prescribed inclination selected in conformity
with the contrast C, the median .gamma.m in the substantial range
.gamma. of compensation of the .gamma. curve (the density
corresponding to the median "128" of the output) coincides with the
threshold density Nth to be designated as illustrated in FIG. 18.
In the case of the data illustrated in the diagram, the luminous
energy of the lamp is set at a value smaller than usual.
When the decision to be made at Step #75 finds the method of
setting the density to be automatic, the threshold density Nth is
selected based on the histogram of density obtained in advance and
the luminous energy of the lamp is set so that the median ym of the
range of density .gamma. of the .gamma. curve coincides with the
selected threshold density Nth (Step #77). The automatic setting of
density is effective only when the image of a given original
document consists mainly of characters. Generally in this case, the
histogram has peaks P1, P2 in each of the character part and the
background part as illustrated in FIG. 19. The threshold density
Nth is defined as the median of the values of density which
correspond to the peaks P1, P2 of each of the character part and
the background part.
Incidentally at Step #76 mentioned above, the switch of the .gamma.
curve may be conceived besides the change of the luminous energy of
the lamp as a method for altering the threshold density Nth
proportionately to the value of density designated by an operator.
The adjustment of density by the luminous energy of the lamp proves
advantageous, however, in due consideration of the optical noise
and the electrical noise in the image sensor 10. For the purpose of
adjusting the density, therefore, it is desirable to utilize the
luminous energy of the lamp within the range in which this luminous
energy allows the adjustment to advantage and then rely on the
switch of the .gamma. curve outside the range.
FIG. 12 is a flow chart of the attribute designation processing to
be performed at Step #6 in the flow of processing shown in FIG.
5.
First, the status is set at "WAIT" (Step #81) and the designation
is checked to find whether or not it is correct (Step #82).
When the designation is not correct as when an area outside the
range set for reading is designated or when an error exists in the
sequence of designation of coordinates, for example, the flow of
processing moves to Step #85 and an error code is transmitted to
the host device.
When the designation is correct, the attribute data write
processing for writing the designation attribute data a0, a1, and
a2 in the attribute memory 30 is executed (Step #83) and the status
is set to "READY" (Step #84).
FIG. 13 is a flow chart illustrating the attribute data write
processing to be performed at Step #83 in the flow of processing
shown in FIG. 12.
At Step #91, the designation from the host device is checked to
find the kind of designation and, depending on the kind thus found,
the various processings of Steps #92 to #98 are executed.
When automatic discrimination of the binarization attribute is
designated, the designated attribute data a0 with respect to the
designated region E is set at "0" at Step #92. When the
binarization attribute has been designated in advance, the
designation attribute data a1 is set at "1" with respect to the
designated region E at Step #93.
When the positivity of image is designated, namely when the
nonnecessity of the white-black conversion is designated, the
designation attribute data a1 is set at "0" with respect to the
designated region E at Step #94. In contrast thereto, when the
negativity is designated, namely the necessity of the white-black
conversion is designated, the designated attribute data a1 is set
at "1" with respect to the designated region E at Step #95.
When the trimming is designated, the designation attribute data a2
is set at "1" with respect other region than the designated region
E at Step #96. When the masking is designated, the designation
attribute data a2 is set at "1" with respect to the designated
region E at Step #97. When the cancellation of trimming-masking
processing is designated, the designation attribute data a2 is
reset at "0" with respect to other region than the designated
region E at Step #98.
FIG. 14 is a flow chart illustrating the output data designation
processing to be performed in the flow of processing at Step #7
shown in FIG. 5.
In this routine, first the output data DATA is checked to find the
kind of data at Step #101 and, depending on the kind of data so
found, the processings at Steps #102 to #104 are executed.
When the outputs exclusively of the image data VIDEO 0 to 7 are
selected, the output control data C0 and C1 are both set at "0" at
Step #102. When the outputs exclusively of the attribute data a4 to
a0 are selected, the processing of Step #103 is executed and the
output control data C0 is set at "1" and the output control data C1
is set at "0."
When the outputs of both the image data VIDEO 0 to 7 and the
attribute data a4 to a0 are selected, the output control data C0 is
set at "0" and the output control data C1 is set at "1" at Step
#104.
In the embodiment described above, since the .gamma. conversion is
effected by preparing a histogram indicative of the density
distribution of an image on a given original document and, based on
this histogram, switching the conversion tables T1 to T3 in
conformity with the contrast C, ideal repeatability of
pseudo-gradient enough to permit production of an image enjoying an
improved quality is attained without reference to the overall
status of density of the image as to lightness or darkness.
In the embodiment described above, the distinction of an image of
characters on a given original document can be exalted without
reference as to the density of characters because the threshold
density Nth of the simple binarization is selected based on the
histogram indicative of the density distribution of an image on an
original document and the adjustment of density is effected by
setting the luminous energy of the lamp during the automatic
setting of density.
In the embodiment described above, the contents of the image
processing can be suitably selected to suit the purpose for which
the produced image is used and such factors as the construction of
each of the component parts of the image reader 1 and the timing of
the operation of the image reader 1 can be altered variously to
fulfill the spirit of this invention.
The embodiment described above represents a case of using an image
reader 1 which is so adapted as to operate by effecting
photoelectric conversion of a scanning light reflected on an
original document. This invention can be applied likewise to a film
scanner adapted to operate by effecting the photoelectric
conversion of a scanning light which has passed through an original
document.
Further, the embodiment described above represents a case of using
a ROM for preparatorily storing therein three .gamma. curves as a
plurality of sets of gradient compensation data. It is permissible
to have one .gamma. curve stored preparatorily in the ROM and,
immediately prior to compensation of gradient, have the .gamma.
curve in the ROM revised in conformity with the contrast of a given
image and put to use for the gradient compensation.
* * * * *